My research program interlinks testing empirical questions in development economics with careful field research. A cross-cutting theme of my research has been attention to the consequences of survey design and a West African focus (Burkina Faso, Mali, Nigeria) where my experience in the region allows me to better adapt my research questions through careful questionnaire design to local institutions and policy issues. Below are a few of my research projects have permitted exploration of survey design issues which are summarized below.
Survey of Household Welfare and Labor in Tanzania
(with Elena Bardasi (World Bank), Kathleen Beegle (World Bank), Pieter Serneels (University of East Anglia), Economic Development Initiatives, TZ)
The quality of labor market statistics is critical for assessing and understanding poverty and living conditions in low-income countries. In low-income countries, especially in Sub-Saharan Africa, much labor is engaged in agriculture (especially in subsistence agriculture), often contributed by family members. Moreover, for women, much labor is allocated to the so-called invisible household economy—that is, in fetching water and wood, carrying out domestic tasks, and providing care to children, the elderly, and persons with disabilities. Employment spans simultaneous (and largely seasonal) activities, paid and unpaid, within and outside the System of National Accounts (SNA). For these reasons, the typical indicators of labor force participation (for example, the employment-to-population ratio, the unemployment rate, main occupation and sector of activity) derived from the standard questions about the “main activity” are generally inappropriate to capture employment patterns such as these which tend to be significantly more complex. The objective of the experiment on labor statistics is to understand the implications of alternative survey methods to measure labor. This work seeks to improve the quality of labor market statistics and the information base for analytical work on employment, with a specific focus on informal sector and female employment. These improvements will include, among other things, better measuring labor force participation, the nature of work in terms of type and intensity (particularly work that occurs in household enterprises and farms), changing patterns in employment over time, and nuanced changes in labor market activity that could otherwise be missed in existing data collection instruments.
The Survey of Welfare and Labor in Tanzania is a randomized experiment which evaluates differences in survey design on consumption and labor statistics. The labor component of this project will estimate the effects of differences in two characteristics of labor modules on differences in labor market statistics. We randomly allocate questionnaire types (a shorter rapid assessment versus a longer LSMS style questionnaire) and respondents (proxy respondents versus self-reporting) to estimate the treatment effects of different survey designs. Specifically, we will investigate differences in labor force participation rates, earnings and hours worked to evaluate differences in survey design. The randomized experiment permits us to control for observable household characteristics (income levels, household composition, education, gender) and unobservable characteristics (power within the household, social norms) that may confound comparisons of labor statistics between two independently administered surveys.
Household Definition Experiments in Mali (with Lori Beaman (Northwestern))
Household definitions used in multi-topic household surveys vary between surveys but have potentially significant implications for household composition, production, and poverty statistics. Standard definitions of the household usually include some intersection of keywords relating to residency requirements, common food consumption, and intermingling of income or production decisions. Despite best practices intending to standardize the definition of the household, it is unclear which types of definitions or which intersections of keywords in a definition result in different household compositions. This paper conducts a randomized survey experiment of four different household definitions in Mali to examine the implications for household-level statistics. This approach permits analysis of the trade-offs between alternative definition types. We find that additional keywords in definitions increase rather than decrease household size and significantly alters household composition. Definitions emphasizing common consumption or joint production increase estimates of the levels of household assets and consumption statistics, but not on per adult equivalency asset and consumption statistics, relative to open-ended definitions of the household. In contrast, definition type did not affect production statistics in levels, though we observe significant differences in per adult equivalency terms. Our findings suggest that variations in household definition have implications for measuring household welfare and production.
Integrated Surveys on Agriculture (with Gbemisola Oseni)
With Gbemisola Oseni, members of the LSMS team, and the National Bureau of Statistics in Nigeria, the General Household of Survey was redesigned in 2010 to include an innovative approach to measuring agricultural acitivities in the context of a household multi-topic survey. The GHS is a cross-sectional survey of 22,000 households throughout the country. The panel component (GHS-Panel) is now being applied to 5,000 households of the GHS and covers multiple agricultural activities. The focus of this panel component is to improve data from the agriculture sector and link this to other facets of household behavior and characteristics. The GHS-Panel drew heavily on the HNLSS and the NASS to create a new survey instrument and method to shed light on the role of agriculture in households’ economic wellbeing. The NBS implemented the first stage (Post Planting) of the first wave of the GHS-Panel in 2010. This panel is a subset of the full GHS (or GHS-Cross Section) that will be finished in 2011.) It is envisaged that the GHS-Panel will be carried out every two years while the GHS-Cross Section will be carried out annually. These studies are funded by the Bill and Melinda Gates Foundation and are publicly available through the LSMS website.